Techniques as described below are known as techniques for monitoring whether a fault occurs in a computer.
For example, a system in which a management server collects information of an apparatus to be monitoredby installing a program for gathering information on an OS (Operating System) operating in the apparatus is known. In this system, the program in a memory of the management server regularly performs polling in order to gather information for the apparatus to be monitored. The program in the memory of the apparatus to be monitored transmits information on a state of the apparatus and information on performance of the apparatus to the management server when the polling from the management server is performed.
However, in this system, because the program installed into the apparatus to be monitored depends on the OS and the program is updated whenever the OS is revised or the like, it is laborious. Moreover, in this system, because the program is resident in the memory of the apparatus to be monitored, there is a problem that a load is applied to the apparatus. Furthermore, in a specific virtual environment, it is difficult to obtain physical information due to virtualization, and information obtained from a host OS can be limited depending on a policy of the OS vendor.
On the other hand, there is a system (hereinafter referred to as an agentless monitoring system) which can monitor without installing the program for gathering information into the apparatus to be monitored. In this system, a j obmonitoringprogram in a job monitoring server reads a log from a job log file of a server to be monitored based on environment setting information, determines a state of a job to be monitored, and outputs the determination result to an intermediate log file when it comes a predetermined time. Furthermore, a log transmission program in the job monitoring server reads an intermediate log from the intermediate log file at predetermined time intervals, and transmits the intermediate log to a remote monitoring server through communication lines.
The agentless monitoring system can solve a problem that occurs in a system other than the agentless monitoring system. However, the agentless monitoring system described above monitors a state of a job in the beginning or the end of the job, but cannot monitor hardware. Moreover, because the log transmission program performs processing at the predetermined time intervals, it is impossible to monitor in real time.
As described above, it is impossible to monitor hardware in an agentless monitoring system in real time by the conventional techniques.
Patent Document 1: International Publication Pamphlet No. WO 2009/144969
Patent Document 2: Japanese Laid-open Patent Publication No. 2011-159011